Abstract The far‐infrared (>16 μm) spectrum plays a crucial role in Earth's radiation budget but was rarely observed from space before 2024. Two Polar Radiant Energy in the Far Infrared Experiment (PREFIRE) satellites, PREFIRE2 (SAT‐2) and PREFIRE1 (SAT‐1), were launched in 2024. They measure spectrally resolved radiances from 5 to 53 μm at 0.84 μm spectral resolution. Deriving spectral outgoing longwave radiation (OLR) and tracking its changes relative to other polar climate factors are key focuses of the PREFIRE mission. We describe a set of algorithms developed to estimate spectral OLR from the PREFIRE L1B‐calibrated radiances. Spectral angular distribution models were developed to invert spectral radiance to spectral flux. For the spectral region not covered by the PREFIRE measurements, a principal‐component‐based linear regression scheme is used to estimate the spectral flux. When the clear‐sky spectral OLRs derived from the PREFIRE SAT‐2 measurements are compared with counterparts derived from in situ measurements, the mean differences in broadband and far‐IR OLR are 7.5 and 3.4 Wm −2 , respectively. The results are also evaluated against collocated broadband and other spectral satellite measurements. The difference (mean ± standard deviation) between the collocated PREFIRE SAT‐2 and CERES broadband OLR is 2.8 ± 5.7 and 2.7 ± 6.5 Wm −2 for clear‐sky and cloudy‐sky footprints, respectively. Similar agreements can be seen from the comparisons with spectral OLR derived from Atmospheric Infrared Sounder and Cross‐track Infrared Sounder measurements. In general, spectral OLR derived from SAT‐2 has better agreement with other collocated measurements than that from SAT‐1, which can be attributed to the higher noisy performance of the SAT‐1 instrument.
Chen et al. (Mon,) studied this question.